Anais De XXXVIII Simpósio Brasileiro De Telecomunicações E Processamento De Sinais 2020
DOI: 10.14209/sbrt.2020.1570646347
|View full text |Cite
|
Sign up to set email alerts
|

K-means clustering for mitigation of nonlinear phase noise in digital coherent optical systems using 16-QAM modulation format

Abstract: This work analyzes the use of the K-means clustering algorithm to mitigate nonlinear phase noise in singlespan coherent systems, such as long-reach passive optical networks (LR-PONs). Simulations revealed that for a 100-km LR-PON employing 16-ary quadrature amplitude modulation (QAM) and considering a 1:64 splitting ratio, the adoption of K-means with K-means++ initialization achieves an optimum bit error ratio (BER) of 6.3 • 10 −4 , whereas employing maximum likelihood, 10 −3 is obtained. We also show that in… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
1
0
1

Year Published

2021
2021
2022
2022

Publication Types

Select...
2
1
1

Relationship

2
2

Authors

Journals

citations
Cited by 4 publications
(2 citation statements)
references
References 24 publications
0
1
0
1
Order By: Relevance
“…Regarding the ANN, a single hidden layer composed of s 2 = 16 neurons. It is also worth noting that for K-means we employed K-means++ as initialization method as this reduces the amount of iterations and reduces the probability for erroneous classification [22]. As can be observed, at low power levels, all the classification methods converge to the BER values of ML but, as the power increases, the different machine learning algorithms outperform ML.…”
Section: Comparison With Other Supervised Classification Techniquesmentioning
confidence: 99%
“…Regarding the ANN, a single hidden layer composed of s 2 = 16 neurons. It is also worth noting that for K-means we employed K-means++ as initialization method as this reduces the amount of iterations and reduces the probability for erroneous classification [22]. As can be observed, at low power levels, all the classification methods converge to the BER values of ML but, as the power increases, the different machine learning algorithms outperform ML.…”
Section: Comparison With Other Supervised Classification Techniquesmentioning
confidence: 99%
“…Assim, no que diz respeito às redes de telecomunicações, as pesquisas relacionadas à AI e ML são aplicáveis nas áreas de transmissão, roteamento e gerenciamento [13]. Nas redes OTN, as técnicas de AI e ML são empregadas nos transponderes ópticos para a caracterização de amplitude e ruído de fase [13], no controle dos amplificadores ópticos [13], na identificação da degradação dos sinais devido aos efeitos lineares de CD e PMD [13], no monitoramento da relação sinal ruído óptica [13], na mitigação dos efeitos não lineares nos receptores [14], [15] e na estimativa da qualidade de transmissão óptica [13]. Além disso, o AI e ML são utilizados nas redes OTN para a classificação automática de modulações [16]- [21], possibilitando o ajuste dinâmico, de tais redes, à qualidade do meio de transmissão, sem a interferência humana, por meio da seleção de taxas de transmissão adequadas ou transmissões em única ou multiportadoras.…”
Section: Introductionunclassified